🚨🚨Super early Chrome deprecation data analysis! 🚨🚨 Data from tens of millions of impressions since the 1% deprecation trial came into effect on Jan 4. First, we actually aren't seeing a full 1% deprecation yet - we think part of this is just timing as well as data discrepancies on our side that we're working through, but we still have enough data to get some rough view of things. Uncookied Chrome users appear to be monetizing about 30% worse than those with cookies. That seems bad, but it's far better than the 60% worse performance of Safari users. Considering that this is just the beginning we think this is a really positive result. We also think much of that 30% gap can be made up this year with continued investment in Privacy Sandbox, ID solutions and more. Our early analysis also shows that Prebid demand (i.e., non-Google demand) is performing better than Google demand in the deprecated user group. We think this is because Google is all-in on Privacy Sandbox, whereas most other SSPs are supporting Sandbox & more. All of this is crazy early data that is likely wrong in some way, and as a side note all of this is display-only as video monetization has different challenges right now. This will all change dramatically in the coming weeks and months but I wanted to share! Bonus data point! We found that Chrome iOS users perform 15% better than Safari iOS users even though they are identically unaddressable (since both run on Apple Webkit). We think this is because of SPO - some buyers ignore Safari users entirely, but don't ignore Chrome. Advertisers! If you're interested in testing how your campaigns will work in the cookieless future, reach out to me. We have fully end-to-end testing available across several DSPs and SSPs.
Premesh Purayil
6mo
Super interesting thanks Paul Bannister . Makes me wonder if the drop in monetization isn't as drastic since it's not the same scene as when Safari did this. Years later we are far more prepared with additional identifiers , id mapping solutions ( probabilistic matched id based solutions had a couple years to map and store data to prepare for this) and first party data. Basically less of a drop since the lights are not really being turned off... Just dimmed I'm curious how much of this changes as time goes by without more change. Mapped IDs in particular seem like they have a shorter shelf life especially on mobile as devices get replaced more often
Paul Bell
6mo
"some buyers ignore Safari users entirely, but don't ignore Chrome." Agreed. Some demand platforms block Safari and have been for some time. That said, we're beginning to see more and more platforms open up.
David Loschiavo
6mo
Your last paragraph is key here... there are plenty of really smart buyside people who are on top of every change, but much of the buyside might not even be aware of the full extent of what's been changed, and have not yet changed their behavior accordingly. They may just be attributing drops in performance to seasonality rather than 3PC deprecation, and that's if they noticed at all -- it is only 1%. And for anyone assuming the buyside is completely on top of this, anyone who dealt with industry wide deprecations (e.g. flash creatives anyone?) knows that some of the largest agencies are not going to address it until the campaign traffic falls to zero. So yes, definitely have to get that buyside feedback.
🚀 Alex Lloro
6mo
Thank you so much for sharing this information. I can't entirely agree with your view as a positive result. Obviously, 30% is much better than 60%, but the market share of Chrome compared to Safari is around 3-3.5 times higher, so 30% is going to have a going to have a more significant impact on the bottom line of advertisers.
Mahesh Narayanan
6mo
Much appreciate your sharing these early results Paul Bannister. Helps all of us learn and deal with this supposedly cataclysmic event. Would appreciate if you can answer a question from this statement in your post : "Uncookied Chrome users appear to be monetizing about 30% worse than those with cookies. " Does this mean, uncookied chrome users converted/performed (i.e. purchased or filled a form etc.) 30% worse than cookied ones?
Siyun F.
6mo
Super interesting and thanks for sharing. My gut guess was 25%-50% worse, unidentifiable vs. cookies. Depending on what tier of publishers in question. Not too far off, I guess :) Definitely will follow closely and see how this shape out. I agree that it will only get better from here. The reason being both buyers and sellers are more prepared, and do have alternatives, on Chrome compared to the case of iOS.
Ufuk Akçay
6mo
The fact that Chrome iOS users are performing 15% better than Safari users is particularly intriguing. It seems to challenge the notion that 'no apple is sweeter than Apple.😄 Jokes aside, this is a great opportunity for advertisers who want more insight into this area. Much appreciate Paul Bannister
Toni Chumillas
6mo
Thanks Paul Bannister . Ver interesting data, in particular with regards to the opportunities arising for solutions that continue investing in Privacy Sandbox and non-cookie ID targeting.
COO
6moThank you Paul Bannister . Definitely great early insight! Question regarding the revenue calculation in both groups: did you look at delivered impressions only, or total bid request that may or may not end up in impression? Former would only consider the cpm impact, while the latter would also look at fill rate, which is important for overal publisher revenue.